ElectroPulse
Limp Gawd
- Joined
- Aug 19, 2013
- Messages
- 129
Hello, all!
A couple of days ago I started studying for the CCENT/CCNA. I've been screwing with Packet Tracer (such a friggin' cool tool!) doing some "what if?" and "how does this work?" experimentation, trying to answer some questions I've been coming up with.
My latest question was "Well, I've read that your MAC address doesn't make it past your router, so how does the layer 2 header change when a frame is sent onto the internet?" I set up a simple topology with a computer1 <-> switch1 <-> router1 <-> router2 <-> switch2 <-> computer2, and watched the frames go through. I pinged from computer1 to computer2, and watched the source and destination MAC addresses change from computer1's interface and router1's MAC (respectively) to router1's WAN interface and router2's WAN interface, then to router1's internal interface and computer2's interface. Question answered.
However, this got me thinking... Routers operate at layer 3, so is there even a functional purpose in updating the MACs in the layer 2 header when being transferred between the routers? Then, this question lead to the question that brought me to create this thread: Why are MAC addresses even used within a LAN? I know most switches operate at layer 2, so MAC addresses are needed, but why don't all switches operate at layer 3 to eliminate this need? I can understand the need for a burned-in address being assigned to interfaces as a unique, unchangeable (MAC spoofing aside) identifier for purposes of DHCP and stuff, but why is it used for transferring data between computers? It feels redundant having two source and destination addresses (MAC and IP) on each frame, when simply an IP address should suffice (at least it appears that way) if everything operated on layer 3.
Anyway, figured I'd ask here to find out the reason... I'm trying to wrap my head around as much networking-related stuff as I can
Thanks!
ElectroPulse
A couple of days ago I started studying for the CCENT/CCNA. I've been screwing with Packet Tracer (such a friggin' cool tool!) doing some "what if?" and "how does this work?" experimentation, trying to answer some questions I've been coming up with.
My latest question was "Well, I've read that your MAC address doesn't make it past your router, so how does the layer 2 header change when a frame is sent onto the internet?" I set up a simple topology with a computer1 <-> switch1 <-> router1 <-> router2 <-> switch2 <-> computer2, and watched the frames go through. I pinged from computer1 to computer2, and watched the source and destination MAC addresses change from computer1's interface and router1's MAC (respectively) to router1's WAN interface and router2's WAN interface, then to router1's internal interface and computer2's interface. Question answered.
However, this got me thinking... Routers operate at layer 3, so is there even a functional purpose in updating the MACs in the layer 2 header when being transferred between the routers? Then, this question lead to the question that brought me to create this thread: Why are MAC addresses even used within a LAN? I know most switches operate at layer 2, so MAC addresses are needed, but why don't all switches operate at layer 3 to eliminate this need? I can understand the need for a burned-in address being assigned to interfaces as a unique, unchangeable (MAC spoofing aside) identifier for purposes of DHCP and stuff, but why is it used for transferring data between computers? It feels redundant having two source and destination addresses (MAC and IP) on each frame, when simply an IP address should suffice (at least it appears that way) if everything operated on layer 3.
Anyway, figured I'd ask here to find out the reason... I'm trying to wrap my head around as much networking-related stuff as I can
Thanks!
ElectroPulse